To reduce this thermal gradient effect, the size of the probe window could be minimized. However, the window needs to be sufficiently large to acquire X-rays scattered at large angles. Therefore, ...

PolarGrad (Polar Gradient methods; Lau et al., 2025) is a class of matrix-gradient optimizers based on the concept of gradient-anisotropy preconditioning in optimization. It has close relation to Muon ...